|Home | About | Journals | Submit | Contact Us | Français|
Oxygen, an essential nutrient, is sensed by a multiple of cellular pathways that facilitate the responses to and survival of oxygen deprivation. The Caenorhabditis elegans embryo exposed to severe oxygen deprivation (anoxia) enters a state of suspended animation in which cell cycle progression reversibly arrests at specific stages. The mechanisms regulating interphase, prophase, or metaphase arrest in response to anoxia are not completely understood. Characteristics of arrested prophase blastomeres and oocytes are the alignment of condensed chromosomes at the nuclear periphery and an arrest of nuclear envelope breakdown. Notably, anoxia-induced prophase arrest is suppressed in mutant embryos lacking nucleoporin NPP-16/NUP50 function, indicating that this nucleoporin plays an important role in prophase arrest in wild-type embryos. Although the inactive form of cyclin-dependent kinase (CDK-1) is detected in wild-type–arrested prophase blastomeres, the inactive state is not detected in the anoxia exposed npp-16 mutant. Furthermore, we found that CDK-1 localizes near chromosomes in anoxia-exposed embryos. These data support the notion that NPP-16 and CDK-1 function to arrest prophase blastomeres in C. elegans embryos. The anoxia-induced shift of cells from an actively dividing state to an arrested state reveals a previously uncharacterized prophase checkpoint in the C. elegans embryo.
Cells have adapted mechanisms to regulate or arrest cell division in response to chemical agents causing DNA damage or microtubule depolymerization. The experimental exposure of cells to these environments (e.g., UV radiation, nocodazole, benomyl) was vital for the identification and understanding of cell cycle checkpoints (Hartwell and Weinert, 1989 ; Nurse et al., 1998 ). Oxygen deprivation, on the other hand, is a natural stressor that many organisms may encounter during growth and development. It was previously observed in several experimental systems, including Caenorhabditis elegans, Drosophila, zebrafish embryos, and mammalian embryonic fibroblasts in culture, that oxygen deprivation induces a reversible cell cycle arrest. Blastomeres of C. elegans and Drosophila melanogaster embryos exposed to anoxia arrest during interphase and all stages of mitosis except anaphase (Foe and Alberts, 1985 ; Padilla et al., 2002 ). In C. elegans, this induced arrest may extend for more than 3 d in the absence of oxygen. Most remarkably, embryonic development proceeds normally after reexposure to air, and no long-term effects on the embryos can be observed; nematodes develop into adults that are morphologically indistinguishable from wild type and produce normal offspring (Padilla et al., 2002 ). D. melanogaster embryos exposed to hypoxia arrest in interphase and the metaphase stage of mitosis (DiGregorio et al., 2001 ; Douglas et al., 2001 ). In vertebrate cells, blastomeres of zebrafish embryos exposed to anoxia arrest during interphase (S and G2 phases; Padilla and Roth, 2001 ) and some cultured mammalian cells exposed to hypoxia arrest in the G1 stage of the cell cycle (Green et al., 2001 ). In all of the determined cases, the arrested cells resume cell cycle progression upon reexposure to air (normoxia). Intriguingly, anoxia-induced suspended animation and embryonic cell cycle arrest occur in such a tightly controlled manner that the arrest is maintained for several days, yet the precise specification of the diverse cell types is not lost. That is, anoxia-induced arrest of embryonic development does not alter the developmental trajectory of cells and tissues. The molecular basis for these defined cell cycle arrests and the control of resumption of the developmental program is poorly understood. It is unclear how a reduction in oxygen signals blastomeres at various stages of cell division to reversibly arrest at specific stages of the cell cycle. Understanding the phenomenon of anoxia-induced suspended animation thus may contribute to a better understanding of how environmental stress influences cell cycle progression and cell cycle checkpoints in developing embryos.
The size and transparency of C. elegans embryos make them an excellent model for observing cellular structures and cell division during development (Moore et al., 1999 ; Oegema et al., 2001 ). Examination of blastomeres from embryos exposed to anoxia indicates that significant alterations in response to oxygen deprivation occur within the nucleus. A cellular hallmark of anoxia-induced arrested blastomeres is the appearance of condensed chromatin that becomes associated with the nuclear periphery; this phenotype is observed in both interphase and prophase blastomeres (Hajeri et al., 2005 ). A characteristic of metaphase blastomeres of embryos exposed to anoxia is the formation of nucleoporin aggregates surrounding the chromosomes, which suggests that the nuclear pore complex (NPC) breakdown is altered under this condition in comparison to untreated embryos at a similar stage of the cell cycle (Hajeri et al., 2005 ). Whether chromatin association with the nuclear periphery and nucleoporin aggregation are essential components of anoxia-induced cell cycle arrest is currently unknown.
Cell cycle progression is controlled by various checkpoints, and evidence from C. elegans and D. melanogaster suggests that anoxia/hypoxia specifically triggers the spindle checkpoint to maintain metaphase arrest. In yeast, mutations in the spindle checkpoint genes lead to sensitivity to microtubule-depolymerizing agents, resulting in a failure to arrest in metaphase in response to spindle damage (Hardwick et al., 1999 ), and mutations in the essential spindle checkpoint genes in mammals may be associated with colorectal and brain tumors (Cahill et al., 1998 ; Reis et al., 2001 ). Exposure to anoxia or microtubule-depolymerizing agents in C. elegans or D. melanogaster embryos deficient for spindle checkpoint function leads to the inability to arrest at metaphase and a chromosome segregation phenotype (Nystul et al., 2003 ; Fischer et al., 2004 ; Encalada et al., 2005 ; Pandey et al., 2007 ; Hajeri et al., 2008 ). Furthermore, C. elegans wild-type embryos exposed to anoxia contain metaphase blastomeres with a reduction in spindle and astral microtubules (Hajeri et al., 2005 ). Experiments in Drosophila show that the process of chromosome attachment to the mitotic spindle might be sensitive to oxygen deprivation, suggesting that the spindle checkpoint is activated by free kinetochores (Pandey et al., 2007 ). Together, these findings are significant because they document that the spindle checkpoint genes are activated in response to the stress of oxygen deprivation in the embryo and that this process is conserved between C. elegans and Drosophila.
Although anoxia-induced metaphase arrest triggered by the spindle checkpoint function is now well established in C. elegans, a different mechanism and checkpoint are responsible for anoxia-induced prophase arrest. Interestingly, the majority of arrested mitotic blastomeres are in either prophase or metaphase, and there are approximately an equal number of arrested prophase blastomeres as there are arrested metaphase blastomeres suggesting that proper arrest of a prophase blastomere is equally as important as arrest of a metaphase blastomere (Padilla et al., 2002 ). The occurrence of different arrest scenarios suggests a dependence upon the blastomeres' position in the cell cycle at the time of exposure to anoxia and the execution by different molecular mechanisms. To begin to understand prophase arrest induced by anoxia, we considered the regulation of cellular events required for the prophase-to-prometaphase transition. Nuclear envelope breakdown (NEBD) is a significant cellular event required for the progression from prophase to prometaphase and a final commitment to mitotic progression (Margalit et al., 2005 ; Anderson and Hetzer, 2008 ). To monitor cell cycle progression under normoxia and anoxia conditions, we conducted live cell imaging and found that an early response to anoxia is chromosome association with the inner nuclear membrane and inhibition of NEBD. The phenomenon of chromosome association with the nuclear periphery was also observed in oocytes of hermaphrodites exposed to anoxia as well as oocytes that are in a quiescent state due to sperm depletion (aged hermaphrodites) or disruption of fertilization (fog-2 mutants). Using RNA interference (RNAi) and genetic mutant analysis, we determined that npp-16, a nonessential component of the NPC, is required for anoxia-induced prophase arrest. Furthermore, we found that in arrested prophase blastomeres of control embryos the inactive state of cyclin dependent kinase (CDK-1) is present, whereas the inactive state of CDK-1 was not detected in arrested prophase blastomeres of the npp-16 mutant embryo. These data support the idea that NPP-16 and CDK-1 function in concert to arrest prophase blastomeres in response to anoxia. Given that oxygen deprivation plays a central role in resistance of solid tumor cells to radiation and chemotherapy treatments, these studies provide novel evidence of how oxygen deprivation influences cell division and reveal mechanistic insight into a poorly understood cell cycle arrest position during mitosis–late prophase arrest.
The C. elegans wild-type N2 Bristol strain was raised on NGM plates seeded with Escherichia coli (OP50) as previously described (Sulston, 1988 ). For all experiments the nematodes were maintained at 20°C. The following strains were obtained from Caenorhabditis elegans Genetics Center and raised under similar conditions as the wild-type strain: TH32 (tbg-1::GFP; pie-1::GFP::H2B), RB1534 (npp-16(ok1839)), CB4108 (fog-2(q71)). For some assays the BN68 strain (pie-1::GFP::npp-5; pie-1::mCherry::his-58), obtained from the Askjaer lab (CABD, CSIC–Universidad Pablo de Olavide, Sevilla, Spain), was used to monitor NEBD (Rodenas et al., 2009 ). The Caenorhabditis elegans Gene Knockout Consortium (Oklahoma Medical Research Foundation) produced the npp-16(ok1839) deletion allele (strain RB1534), which has a 1120-base pair deletion resulting in removal of amino acids 175-472 of the 512-amino acid nucleoporin NPP-16 protein. The npp-16(ok1839) was backcrossed three times to wild-type N2 animals to produce the strain PM118 in which all analyses were conducted. Using standard genetic techniques, the TH32 strain was crossed with the npp-16(ok1839) deletion PM118 strain to produce strain PM119 (npp-16(ok1839); tbg-1::GFP; pie-1::GFP::H2B) or the CB4108 strain to produce PM122 (fog-2(q71); tbg-1::GFP; pie-1::GFP::H2B). The genotype of npp-16(ok1839) allele was verified by conducting single worm PCR using npp-16 forward primer (5′-TGACTCATCGAGCCTGAAAA-3′) and npp-16 reverse primer (5′-GAGTCGAACTTCCCAAGCAG-3′).
Live animal imaging analysis was conducted using strains TH32 (tbg-1::GFP; pie-1::GFP::H2B) and PM119 (npp-16(ok1839); tbg-1::GFP; pie-1::GFP::H2B), which are both expressing γ-tubulin::GFP and histone2B::GFP, as previously described (Hajeri et al., 2008 ). These strains allow one to monitor mitotic progression by visualization of centriole location and chromosome condensation based on green fluorescent protein (GFP) fluorescence detection. Nematodes were grown to adulthood on NGM plates seeded with E. coli. Gravid adults were anesthetized (0.5% tricaine, 0.05% tetramisole [Sigma-Aldrich, St. Louis, MO] in M9 buffer) and mounted on 2% agarose pads placed on coverslips (Warner Instruments, Hamden, CT); this method to anesthetize adults slows their movement during microscopic examinations, but does not interfere with ovulation and fertilization. The anesthetized nematodes were coated with a drop of halocarbon oil to prevent desiccation during exposure to a flow of nitrogen gas or air and placed in an enclosed chamber (Harvard Apparatus, South Natick, MA; Leiden closed perfusion microincubator). To monitor anoxia-induced prophase arrest, the chamber was perfused with 100% nitrogen (Air Liquide-Calgaz, Cambridge, MD), and time-lapse images of embryos within the uterus of the adult nematode were collected using a spinning disk confocal microscope (McBain Systems, Simi Valley, CA; see below). To document recovery of arrested prophase blastomeres, the nitrogen flow to the perfusion chamber was discontinued, and embryos were allowed to recover in air and imaged using time-lapse microscopy. The integrated Simple PCI software program (Hamamatsu, Bridgewater, NJ) was used for image analysis and to assess the statistics of anoxia-induced prophase arrest and recovery. To analyze mitotic progression in animals exposed to normoxia conditions, a similar methodology was followed without the nitrogen flow through the chamber. The time-lapsed movies were processed by using ImageJ (NIH; http://rsb.info.nih.gov/ij/) and imported into QuickTime (Apple Computer, Cupertino, CA) for display.
To analyze the nuclear location and movement of bivalent chromosomes in oocytes of animals exposed to stress (anoxia, starvation, sodium azide) or animals with quiescent oocytes (aged animals, fog-2 mutant), time-lapse microscopy of the primary oocytes of TH32 animals (tbg-1::GFP; pie-1::GFP::H2B) and PM122 (fog-2(q71); tbg-1::GFP; pie-1::GFP::H2B) animals was conducted. For anoxia assays, 1-d-old TH32 adults were exposed to 24 h of anoxia using the anaerobic bio-bag type A environmental chamber as previously described (Padilla et al., 2002 ). After anoxia exposure, animals were quickly mounted on a 2% agar pad, and time-lapse microscopy of oocytes was conducted; all image acquisition was performed within 10 min of reexposure to normoxia. Control oocytes were from 1-d-old TH32 animals exposed to a normoxic environment. For analysis of oocytes from starved animals, 1-d-old hermaphrodites were washed four times in M9 buffer to remove E. coli and then placed on an unseeded NGM plate for 12 h. These starved adults held embryos within the uterus, indicating that the lack of food was indeed inducing a physiological response. To analyze oocytes of 1-d-old hermaphrodites with blocked respiratory function, animals were placed in M9 buffer with 10 mM sodium azide for up to 1 h. Quiescent oocytes were analyzed from aged TH32 animals (6 d after L4 molt) and 1-d-old PM122 (fog-2(q71), pie-1::GFP::H2B) animals. For all assays, time-lapse microscopy of primary oocytes was conducted, to examine the localization and movement of bivalent prophase chromosomes within the nucleus, using a spinning disk confocal microscope (McBain Systems) and processed using ImageJ software (NIH) and Adobe Photoshop (San Jose, CA).
For indirect immunofluorescence assays embryos from young adults were collected and processed as previously described (Hajeri et al., 2005 ); with the exception of anti-CDK-1 the method used is as described in Boxem et al. (1999) . The following primary antibodies were used: anti-Phos H3 to detect the phosphorylated (Ser10) form of Histone H3 (Upstate Biotechnology, Lake Placid, NY); mAb414 to detect the nuclear pore complex (Babco, Berkeley, CA); anti-HCP-3 to detect the centromeric histone HCP-3 (Buchwitz et al., 1999 ); YL1/2 antibody to detect microtubules (Amersham Life Science, Little Chalfont, Buckinghamshire, England); and anti-CDK-1P-Tyr15 antibody to detect the inactive form of CDK-1 (Calbiochem, San Diego, CA; Hachet et al., 2007 ) and a polyclonal antibody raised against the C terminus of CDK-1 (a kind gift from A. Golden [National Institute of Diabetes and Digestive and Kidney Diseases, Bethesda, MD] and J. Schumacher [University of Texas M.D. Anderson Cancer Center, Houston, TX]; Boxem et al., 1999 ). All primary antibodies were used in a working dilution of 1:500, except anti-CDK-1P-Tyr15 was used in a working dilution of 1:100 and anti-CDK-1 was a 1:1000 working dilution. Embryos were examined using a Zeiss Axioscope fluorescence microscope (Thornwood, NY). Embryos were examined and images obtained using a spinning disk confocal microscope (McBain Systems) at a consistent temperature of 22°C. Specifications for the confocal microscope were as follows: Zeiss 200M microscope with Apochromat 40×, LCI Plan Neofluar 63×, or Plan Apochromat, 100× objectives, and images were acquired with a Hamamatsu Electron Multiplier CCD camera (model C9100-02). Image acquisition software was Simple PCI (Hamamatsu) and images were processed using ImageJ software (NIH) and Adobe Photoshop (Hajeri et al., 2005 ).
Analysis of the germline was conducted in a similar manner as previously described (Jud et al., 2008 ). Briefly, 1-d-old hermaphrodites were transferred to M9 and washed twice to remove bacteria. Approximately 20 animals were placed on a coverslip and gonads were extruded using a 27-gauge syringe needle, fixed for 20 min in 2% paraformaldehyde in a humid chamber, placed on a poly-lysine–coated slide, freeze-cracked, and fixed in −20°C methanol for 10 min. Slides were incubated with mAb414 for 1 h and secondary antibody for 30 min at room temperature. To stain DNA, DAPI was included in the last rinse before mounting for microscopy analysis. The dissected gonads were examined using both a Zeiss Axioscope fluorescence microscope or a spinning disk confocal microscope (McBain Systems); image acquisition was done using Simple PCI (Hamamatsu) and processed using ImageJ software (NIH) and Adobe Photoshop (Hajeri et al., 2005 ).
To determine if microtubule depolymerization induced prophase arrest and prevented NEBD, the TH32, BN68, or N2 embryos were treated with the microtubule-depolymerizing drug nocodazole (Sigma-Aldrich) as previously described (Moore et al., 1999 ; Nystul et al., 2003 ). Gravid adults were dissected in M9 buffer to release embryos and exposed for 5 min to nocodazole (0.2 mg/ml final concentration) or M9 for controls. Embryos were collected by mouth pipetting, mounted on 2% agar pads, and covered with a coverslip; slight pressure on the coverslip was placed to ensure nocodazole entry into the embryo. Time-lapse microscopy using a spinning disk confocal microscope was conducted with the TH32 or BN68 embryos. Indirect immunofluorescence assays were performed with fixed N2 embryos to assay efficiency of microtubule depolymerization by nocodazole treatment.
For analysis of embryos exposed to the respiratory chain inhibitor, sodium azide, embryos were dissected from gravid adults and placed in M9 buffer or M9 buffer containing sodium azide (10 mM) (Sigma-Aldrich) for 30 or 60 min. After sodium azide treatment, embryos were frozen on dry ice, fixed, and stained with DAPI, Phos-H3, and mAb414 antibodies. Image acquisition and analysis was done as described above. To determine the viability of embryos exposed to sodium azide, embryos were collected after treatment, washed three times in M9 buffer, and then placed on an NGM plate seeded with E. coli. The embryos were maintained at 20°C and assayed for their ability to hatch after 24 h. Three independent assays, with a total of at least 200 embryos, were conducted. Data were evaluated using Student's t test to determine p values.
A synchronous population of TH32 (tbg-1::GFP; pie-1::GFP::H2B) L1 larvae were grown to adulthood on RNAi plates (NGM supplemented with 200 μg/ml ampicillin, 12.5 μg/ml tetracycline, and 2 mM IPTG). The RNAi plates were seeded with the E. coli strain for RNAi of a specified gene of interest. The E. coli strains were developed by the J. Ahringer laboratory and obtained from the Medical Research Council Geneservice (Cambridge, United Kingdom; Kamath et al., 2003 ). As a control, worms were fed the E. coli HT115 strain containing the plasmid L4440 without insert. Adult RNAi-treated animals, which did not show a phenotype under normoxic conditions (npp-14(RNAi), npp-16(RNAi) and npp-18(RNAi)) were exposed to 24 h of anoxia using the anaerobic bio-bag Type A environmental chamber (Fisher Scientific [Pittsburgh, PA]; Padilla et al., 2002 ; Hajeri et al., 2008 ). After anoxia exposure, the adults were quickly transferred to a 2% agar pad and assayed for the presence of prophase-arrested blastomeres and abnormal nuclei by spinning disk confocal microscopy. To determine if RNAi of npp-14, npp-16, or npp-18 affected mitotic progression in normoxic embryos, gravid adults were anesthetized (0.5% tricaine, 0.05% tetramisole in M9) and mounted on 2% agarose pads, and live cell imaging of the embryos in the uterus was performed. Anoxia viability assays for embryos were conducted as previously described (Padilla et al., 2002 ; Nystul et al., 2003 ). Young adults were dissected, and two-cell embryos were collected, placed on NMG plates, and exposed to either normoxic or anoxic environments (anaerobic bio-bag Type A environmental chamber) for 24 h. On reexposure to oxygen, the embryos were given 24 h to recover, and survivors were quantified.
To determine the nuclear permeability of embryos under normoxic and anoxic conditions, the TH32 (tbg-1::GFP; pie-1::GFP::H2B) and PM119 (npp-16(ok1839); tbg-1::GFP; pie-1::GFP::H2B) animals were injected in one syncytial gonad arm with either 10-, 40-, or 70-kDa dextran coupled to tetramethylrhodmine (TMR) isothiocyanate (Sigma-Aldrich) as described by Galy et al. (2003) . The TMR-dextran molecule was diluted to a final concentration of 2 mg/ml in injection buffer (20 mM KPO4, pH 7.5, 3 mM K citrate, 2% polyethylene glycol-6000). Injected worms were incubated for 4 h at 23 or 20°C to allow incorporation of the dextran into newly formed embryos. Adult animals were either directly dissected for imaging of the normoxic embryos or exposed to 24 h of anoxia before dissection, using the anaerobic bio-bag Type A environmental chamber, as previously described (Padilla et al., 2002 ). Nuclear permeability was evaluated by assaying the exclusion or inclusion of fluorescently labeled dextran within the nucleus. The TH32 strain with GFP-labeled chromosomes and centrioles was used to identify mitotic cells. Embryos were analyzed, and images were collected using a Zeiss Mot II Plus microscope, Axiocam camera, and software.
Blastomeres of C. elegans embryos exposed to anoxia arrest during interphase and specific stages of mitosis (primarily metaphase and prophase and not anaphase; Foe and Alberts, 1985 ; Padilla et al., 2002 ). Previously we have shown that arrested prophase blastomeres of embryos exposed to anoxia have fully condensed chromosomes that align at the nuclear periphery (Hajeri et al., 2005 ). To further characterize prophase chromosome alignment with the nuclear periphery and prophase arrest, we conducted live-cell imaging analysis using a C. elegans strain expressing γ-tubulin::GFP and histone2B::GFP fusion proteins (tbg-1::GFP; pie-1::GFP::H2B). This strain, referred to as TH32, was used throughout our study to follow chromosome location within the nucleus and cell cycle progression. Embryos, within the uterus of gravid adult animals placed in a gas flow-through chamber, were visualized using a spinning disk confocal microscope. The animals were exposed to either air (normoxia) or nitrogen gas (anoxia). We found that the condensed chromosomes aligned at the nuclear periphery in prophase blastomeres exposed to anoxia for 8.5 min, demonstrating the fast response to anoxia conditions. Furthermore, the round nucleus morphology was maintained suggesting an arrest of NEBD (Figure 1A, Video 1). On reexposure to air, within ~4 s, the chromosomes move off the nuclear envelope, and the arrested prophase blastomere progresses through mitosis, as evidenced by the observation of NEBD and formation of the metaphase plate (Video 2). We observed that the longer the exposure time to anoxia (24 h anoxia), the longer (>15 min) it took for the chromosomes aligned with the inner nuclear periphery to move off (data not shown). This is consistent with our previous findings that recovery from cell cycle arrest is dependent on anoxia exposure time (Hajeri et al., 2005 ). In control prophase blastomeres of embryos exposed to normoxia, the chromosomes localized without specific arrangements and were dispersed throughout the nucleus (Figure 1A). Next, we used indirect immunofluorescence to detect the centromeric histone CENP-A/HCP-3 and NPCs and found that the centromere structure was not altered in prophase chromosomes that were associated with the nuclear periphery, further suggesting that chromosome structure remains intact in arrested prophase blastomeres (Figure 1B). The ability of anoxia-exposed embryos to maintain chromosome structure is in agreement with our previous finding that these embryos are able to recover from the arrest and progress normally through embryogenesis (Padilla et al., 2002 ).
Because prophase cells are not only present in the embryo but also within the germline, we next investigated the impact of anoxia on the chromosome organization in oocytes. The oocytes in adult hermaphrodites are in diakinesis of prophase I and contain observable bivalent chromosomes (McCarter et al., 1999 ). Previously, we showed that anoxia induces an arrest of oocyte maturation and ovulation in the adult hermaphrodite (Mendenhall et al., 2009 ). We, therefore, wanted to determine if anoxia also affects bivalent chromosome location within the nucleus of oocytes. We analyzed oocytes in adults exposed to 24 h of anoxia (using the anaerobic biobags) or brief periods of anoxia using the N2 gas flow through chamber. In control animals exposed to normoxic conditions, the bivalent chromosomes are found to move throughout the nucleoplasm of oocytes in hermaphrodites (Figure 1C). However, using the TH32 strain, we indeed found that the bivalent chromosomes aligned with the nuclear periphery in arrested oocytes of adult hermaphrodites exposed to anoxia (Figure 1C). This chromosome alignment was observed in all oocytes and not just the primary oocyte (Figure S1). Oocytes of adult animals exposed to anoxia are viable and able to be fertilized after anoxia treatment, as determined by oocyte maturation rate assays conducted in postanoxic hermaphrodites (data not shown). We did find that after longer exposures to anoxia (24 h) it took ~1.5 h for oocytes to recover in air and proceed with ovulation (n = 6). To determine how quickly the chromosomes respond to anoxia we used the N2 gas flow-through chamber and found that chromosomes aligned with the inner nuclear periphery after exposure to 12 min of N2 gas flow. On reexposure to normoxic conditions, the chromosomes within the primary oocyte moved off the nuclear envelope just before maturation and ovulation, whereas the chromosomes in the other oocytes remained associated with the inner nuclear periphery (n = 6 germlines observed). This suggests that the recovery process may be linked with signals involving oocyte maturation and ovulation. Furthermore, these results show that similar to prophase blastomeres, oocytes also contain chromosomes associated with the inner nuclear periphery after exposure to anoxia.
Because anoxia is a stress phenomenon that induces arrest of oocyte fertilization, we next wanted to determine if bivalent chromosome alignment with the nuclear periphery is a characteristic that is also associated with stresses and/or with quiescent oocytes. We found that starving adult hermaphrodites for 12 h, which caused adults to hold embryos in the uterus (data not shown), did not induce bivalent chromosomes to align with the nuclear periphery (Figure 1C). To test whether bivalent chromosome alignment with the nuclear periphery was induced by lack of respiratory function, we treated animals with sodium azide, an inhibitor of cytochrome C in the electron transport chain, which is also known to induce ovulation arrest. Indeed, as shown in Figure 1C, many of the bivalent chromosomes also aligned with the nuclear periphery in animals exposed to sodium azide. This phenotype may be associated with the azide-induced ovulation arrest. We found that bivalent chromosomes also aligned with the nuclear periphery in quiescent oocytes of animals in which ovulation does not or only rarely occurs due to sperm depletion (aged hermaphrodites) or nonfunctional sperm (fog-2(q71) mutant; Figure 1C; Schedl and Kimble, 1988 ; Jud et al., 2008 ). We did not observe any major difference in the chromosome association with the inner nuclear periphery in anoxia-arrested oocytes compared with the quiescent oocytes due to sperm depletion or dysfunction. These data suggest that bivalent chromosome association with the nuclear periphery is not signaled by stress per se but is rather a phenotype associated with stress-induced arrest or quiescence.
To further investigate the mechanism that regulates prophase arrest, we considered two hypotheses: 1) Prophase arrest occurs in response to a reduction in energy, or 2) prophase arrest is regulated by a genetic mechanism, such as a defined cell cycle checkpoint. The first hypothesis is based on the finding that the ATP/ADP ratio is decreased in C. elegans exposed to anoxia (Padilla et al., 2002 ). The second hypothesis is based on the finding that anoxia-induced metaphase arrest requires the spindle checkpoint pathway; therefore, it is plausible to assume that prophase arrest is also controlled by a cell cycle checkpoint (Nystul et al., 2003 ). To address the first hypothesis, we exposed embryos to sodium azide, to reduce ATP production by inhibiting electron transport activity, and examined chromosome location in prophase blastomeres and embryo viability. Indirect immunofluorescence and antibodies to detect the phosphorylated forms of histone H3 (Phos H3) or NPC (mAb414) were used to identify prophase blastomeres (Padilla et al., 2002 ). We found that chromosomes aligned with the nuclear periphery in prophase blastomeres of embryos exposed to sodium azide for 30 or 60 min (Figure 2A), suggesting that a reduction in energy, via sodium azide exposure, is sufficient to initiate chromosome relocalization. However, in mitotic blastomeres of embryos exposed to sodium azide a dramatic increase in abnormal chromosome morphology was observed between 30- (2.38%, n = 304) and 60-min exposure to azide (24.3%; n = 239; Figure 2A, arrow). Furthermore, unlike embryos exposed to anoxia, which can survive up to 3 d in the absence of oxygen (Padilla et al., 2002 ), the embryos exposed to 1 h of sodium azide had a significantly decreased survival rate (Figure 2B), indicating fundamentally different molecular mechanisms of cell cycle arrest.
To test the hypothesis that anoxia-induced prophase arrest is regulated by a genetic mechanism, such as a cell cycle checkpoint, we considered the cell cycle events that must be regulated in the arrested prophase blastomere. The transition from prophase to prometaphase is characterized by chromosome condensation, NEBD, and the initiation of chromosome movement to the equatorial plate; NEBD is considered a commitment to mitosis. The regulation of NEBD is not completely understood, but several studies show that the mechanical tension produced by microtubule polymerization may tear holes in the nuclear envelope, thus initiating NEBD (Beaudouin et al., 2002 ; Salina et al., 2002 ). We have previously shown that microtubules in fact depolymerize in embryos exposed to anoxia (Hajeri et al., 2005 ). Therefore, to determine if microtubule depolymerization generates a signal for chromosomes to align with the nuclear periphery and/or arrest cells at prophase, we exposed embryos to nocodazole. We tested this in three different strains (N2, TH32, and BN68) and stained for microtubules, NPC, and DNA in the N2 strain (Figure S2) or conducted live imaging analysis of the prophase blastomeres in the TH32 strain (data not shown) or BN68 strain (Videos 3 and 4). We did not observe chromosome alignment with the nuclear membrane or prophase arrest in these embryos (Figure S2, Video 4). Embryos exposed to nocodazole were affected because there was a decrease in detection of microtubules (Figure S2), and embryos contained abnormal nuclei due to DNA material outside of the nucleus (Video 4). Our findings are consistent with other reports showing that microtubule polymerization is not essential for NEBD (Encalada et al., 2005 ; Margalit et al., 2005 ).
It is postulated that NEBD is preceded by a stepwise disassembly of NPC and activation of specific kinases, such as cyclin-dependent kinase (CDK1), to phosphorylate proteins associated with the inner nuclear membrane (Margalit et al., 2005 ; Galy et al., 2008 ). The NPC in mammals is composed of about ~30 nucleoporins, several of which are known to have additional functions besides the transport of macromolecules in and out of the nucleus (D'Angelo and Hetzer, 2008 ). In Drosophila and C. elegans embryos and starfish oocytes the disassembly of nucleoporins occurs before the disassembly of inner nuclear membrane–associated proteins (Lee et al., 2000 ; Kiseleva et al., 2001 ; Lenart et al., 2003 ). Therefore, we next tested if nucleoporin function was involved in the signaling of anoxia-induced prophase arrest. In C. elegans there are 24 genes predicted to encode nucleoporins (Galy et al., 2003 ; D'Angelo and Hetzer, 2008 ). Using RNAi to knock down individual NPC components, we found, consistent with previous reports, that most nucleoporin genes serve essential functions and accordingly knockdown resulted in various phenotypes including embryo lethality, sterility, larval lethality, and larval arrest. However, RNAi of the nucleoporin genes npp-14, npp-16, or npp-18 (predicted homologues of mammalian Nup214, Nup50, and Seh1, respectively) did not result in obvious phenotypes (Galy et al., 2003 ); this allowed us to test the hypothesis that nucleoporins are involved in anoxia-induced prophase arrest. We therefore exposed npp-14(RNAi), npp-16(RNAi), and npp-18(RNAi) animals, in the TH32 strain background, to anoxia and analyzed prophase chromosome location and nuclear structure. Indeed, we found that the npp-16(RNAi) embryos exposed to anoxia contained blastomeres with abnormal nuclei (data not shown). These observations were confirmed and further analyzed with an npp-16 deletion mutant, npp-16(ok1839). We found that young npp-16(ok1839) embryos have a reduced viability when exposed to anoxia (53.1% ± 2.1), but develop almost normally in the presence of normoxia (91.6% ± 2.1); N2 control embryos survive anoxia at a rate of 100%. By using indirect immunofluorescence we observed that the npp-16(ok1839) embryos exposed to anoxia contain some prophase blastomeres in which the chromosomes aligned with the nuclear periphery. In addition, the embryos also contained blastomeres with abnormal nuclei, as determined by abnormal chromosome structure (Figure 3, A and B, arrow) and aggregation or abnormal distribution of NPC (Figure 3, A and B, arrowhead). These phenotypes were not observed in the npp-16(ok1839) embryos under normoxia conditions (Figure 3A).
To determine if the abnormal nuclei observed in anoxia exposed npp-16(ok1839) embryos were due to blastomeres that were unable to arrest at a specific position during mitosis, we quantified the mitotic index for young embryos. We determined that the npp-16(ok1839) embryos exposed to anoxia harbored a decreased number of prophase blastomeres and an increase in abnormal nuclei (Figure 4), compared with wild-type control animals under these conditions. The number of metaphase blastomeres was not significantly altered between control and npp-16(ok1839) embryos exposed to anoxia, suggesting that npp-16 function is specific for prophase arrest and not involved in metaphase arrest. There was no significant difference between the mitotic indices of npp-16(ok1839) and control embryos exposed to normoxia (Figure 4). Together, these results suggest that npp-16 function is required for anoxia-induced prophase arrest, and that in the absence of npp-16 function the cell cycle progresses, thus leading to abnormal nuclei.
To further determine if the presence of abnormal nuclei in npp-16(ok1839) embryos exposed to anoxia is due to the inability of prophase blastomeres to arrest, we conducted live cell imaging analysis in both the TH32 strain and in npp-16(ok1839) mutant strain, which was backcrossed into the TH32 strain (Figure 5, Videos 5 and 6). As previously described (Figure 1A), prophase blastomeres from the TH32 control strain exposed to anoxia contain chromosomes that associate with the nuclear envelope (Figure 5A) and upon exposure to air the arrested blastomeres continue to progress through mitosis, as seen by the formation of the metaphase plate and chromosome segregation (Figure 5B). Mitotic progression from prophase to anaphase is normal in npp-16(ok1839) embryos exposed to normoxia (Figure 5C, Video 5). However, the npp-16(ok1839) mutant embryos exposed to anoxia contained prophase blastomeres in which the chromosomes did not completely align with the nuclear periphery (8 of 26 prophase blastomeres) or NEBD occurred, as determined by the eventual loss of a circular nucleus, resulting in blastomeres with abnormal chromosome structure (8 of 26 prophase blastomeres analyzed; Figure 5D, Video 6). However, in some cases (10 of 26) prophase arrest, and chromosome alignment with the nuclear periphery occurred, indicating that the abnormal anoxia arrest phenotype is not 100% penetrant in the npp-16(ok1839) mutant. Together, these data further support the notion that, unlike control embryos, the prophase blastomeres from npp-16(ok1839) animals do not consistently arrest in response to anoxia, pointing to a novel important function of nucleoporins in oxygen-controlled cell cycle progression.
NPC distribution, as detected by the mAb414 mAb, was abnormal in some npp-16(ok1839) blastomeres of embryos exposed to anoxia (Figure 3, arrowhead). Therefore, we considered the possibility that the permeability of the nuclear envelope was compromised in npp-16(ok1839) embryos under these conditions. The permeability of the nuclear envelope can be monitored by using fluorescently labeled dextran molecules (Galy et al., 2003 ; Lenart and Ellenberg, 2006 ); smaller molecules 20 kDa or below are able to diffuse through the NPC; however, if the NPC transport function is compromised, then larger macromolecules (70 kDa) will also pass into the nucleus (Galy et al., 2003 ; Lenart and Ellenberg, 2006 ). To assay nuclear permeability of control and npp-16(ok1839) embryos exposed to either anoxia or normoxia, we used a fluorescently labeled 70-kDa dextran. To assay the nuclear permeability in the context of the respective cell cycle position, we used TH32 (control strain) and npp-16(ok1839) crossed into the TH32 background, which express GFP-tagged γ-tubulin and histone2B. The 70-kDa dextran was excluded from prophase and interphase nuclei of both control and npp-16(ok1839) embryos exposed to either anoxia or normoxia (Figure 6, Table S1). Notably, the 70-kDa dextran was also excluded from nuclei in the abnormal blastomeres of npp-16(ok1839) embryos exposed to anoxia (Figure 6, arrow), indicating that the permeability barrier through the NPC or nuclear envelope remained intact, despite the abnormal NPC distribution associated with npp-16(ok1839) embryos. To determine if the nucleus is more permeable to molecules smaller than 70 kDa, we conducted similar assays using the 40- and 10-kDa dextran molecules. The 40-kDa dextran molecule behaved similarly to the 70-kDa dextran molecule in that dextran was excluded from the prophase and interphase nuclei in both control and npp-16(ok1839) embryos exposed to normoxia or anoxia (data not shown). This data indicates that nucleus permeability was likely not severely compromised in these blastomeres. The 10-kDa dextran molecule is small enough to diffuse into the nucleus, and we did observe diffusion of the 10-kDa dextran molecule in many, but not all, of the nuclei in both control and npp-16(ok1839) blastomeres exposed to anoxia or normoxia (Table S1). Note that in several cases we observed that the dextran molecules (70, 40, and 10 kDa) aggregated within the cytoplasm, which could contribute to the inability to diffuse into the nucleus. However given that the 70- or 40-kDa dextran molecules did not diffuse into the nucleus of npp-16(ok1839) blastomeres exposed to anoxia suggest that the requirement of npp-16 function to arrest prophase blastomeres in response to anoxia is not merely due to abnormal nuclear transport of macromolecules.
To characterize the germ line phenotype in npp-16(ok1839) hermaphrodites, we used the TH32 strain backcrossed into the npp-16(ok1839) strain or stained the gonad, of npp-16(ok1839) hermaphrodites, with mAb414 and DAPI to recognize the NPC and chromosomes, respectively. We observed several phenotypes associated with the npp-16(ok1839) mutation. First, unlike control animals, the npp-16(ok1839) hermaphrodites exposed to normoxia contain oocytes in which the bivalent chromosomes associate with the inner nuclear periphery; this was observed in animals analyzed by live imaging analysis or by immunostaining of the gonad (Figure S3 and Table S2; Figure 7B). Analysis of the npp-16(ok1839) hermaphrodites exposed to anoxia indicates that many of the oocytes had an abnormal nucleus as determined by an abnormal shape or if the membrane, as detected by mAb414, was not round/oblong and appeared to have invaginations or folds (Table S2 and Figure S3; Figure 7B). We also observed phenotypes associated with anoxia exposure (Figure 7). Others have shown, in the wild-type germ line, that mAb414 localizes at cluster regions in germ cell nuclei (Pitt et al., 2000 ). We observe in control germ cells mAb414 localized around the nuclear DNA. However, in both wild-type and npp-16(ok1839) animals the localization of mAb414 is more dispersed in the germ cells within the pachytene region of the gonad (Figure 7B), suggesting that the NPC of these germ cells are perturbed in response to anoxia. Second, in oocytes the mAb414 detection accumulates at the cortical region of the oocyte and is not as intense at the nucleus in comparison to normoxic controls (Figure 7A, cf. oocytes noted by asterisk and arrow). Third, the sperm are always visible in the gonad of anoxia-exposed animals (Figure 7A, region noted as SP); this suggests that ovulation may arrest in a specific manner, but further analysis of the spermatheca would need to be conducted to determine if such is the case. Together, these results suggest that in the oocyte of animals exposed to anoxia, npp-16 is not required for chromosome association with the nuclear periphery but is required for maintenance of nuclear shape. Furthermore, anoxia affects the localization of NPC in the meiotic cells of both wild-type and npp-16(ok1839) animals, further supporting the idea that the NPC has a role in anoxia responses. To determine if aged oocytes were viable in npp-16(ok1839) hermaphrodites depleted of sperm, as determined by the lack of offspring and occasional unfertilized oocyte on NGM plate, the hermaphrodites were mated with adult N2 males; 62.5% of the npp-16(ok1839) hermaphrodites (n = 16) were able to produce offspring in which a large frequency of males were observed. In comparison, 83% of sperm-depleted wild-type hermaphrodites produced offspring (n = 6). Together our data indicates that the gonad of npp-16(ok1839) animals is functional but has some phenotypes associated with the mutation.
Several nucleoporins are known to have transport-independent functions including a possible role in NEBD, kinetochore function, gene regulation, and chromosome segregation (D'Angelo and Hetzer, 2008 ; Capelson and Hetzer, 2009 ). To further test the hypothesis that prophase arrest is controlled genetically and that this control is compromised in npp-16(ok1839) embryos, we conducted a cell biological analysis to assay CDK-1 activity. In yeast and mammalian cells, a critical component for the commitment and progression through mitosis is activation of the Cdk1/cyclin B1 complex (Nurse, 1990 ; Margalit et al., 2005 ). In our hands, the sterility observed in cdk-1(RNAi) hermaphrodites is near 100% (data not shown), and thus it is not straightforward to genetically determine if CDK-1 function is necessary for anoxia-induced prophase arrest in embryos (Burrows et al., 2006 ). Therefore, we took a cell biological approach to investigate CDK-1 function in relation to prophase arrest. We used antiserum raised against the C-terminus of CDK-1 (anti-CDK-1) or a phospho-specific antibody (CDK-1P-Tyr15) to detect the inactive form of CDK-1 (Figure 8). Others have shown, using the anti-CDK-1 antibody, that CDK-1 is localized to the nucleus and cytoplasm in C. elegans embryos (Boxem et al., 1999 ). We found that in normoxic control and npp-16(ok1839) embryos (n = 58 and 42, respectively) the majority of prophase blastomeres contain CDK-1 localized to the cytoplasm and nucleus yet in some cases anti-CDK-1 signal is brighter in the cytoplasm relative to the nucleus (21.6 and 5.9%, respectively) or brighter in the nucleus relative to the cytoplasm (10.8 and 23.5%, respectively). Interestingly, embryos exposed to anoxia contain prophase blastomeres in which CDK-1 is localized on or near the chromosomes associated with the inner nuclear periphery (Figure 8A arrow). In wild-type embryos exposed to anoxia the majority of prophase blastomeres (92.8%, n = 28) contained anti-CDK-1 localized with the chromosomes, whereas this was the case with ~50% (n = 10) of the npp-16(ok1839) prophase blastomeres of embryos exposed to anoxia. The anti-CDK-1 was detected on or near chromosomes within abnormal blastomeres of npp-16(ok1839) embryos exposed to anoxia (40%, n = 30). The association with chromosomes was not observed in embryos exposed to normoxia indicating that CDK-1 localization is altered in response to anoxia.
During interphase, Cdk1 is phosphorylated and in an inactive form, and its dephosphorylation during mitosis activates the complex and leads to Cdk1-catalyzed phosphorylation of inner nuclear membrane proteins which is required for NEBD in mammalian cells (Heald and McKeon, 1990 ; Margalit et al., 2005 ). To determine the localization pattern of the inactive form of CDK-1 in C. elegans exposed to anoxia, we used the phospho-specific antibody (CDK-1P-Tyr15) in wild-type and npp-16(ok1839) embryos (De Souza et al., 2000 ; Hachet et al., 2007 ). Similarly to what others have shown in the C. elegans zygote (Hachet et al., 2007 ), inactive CDK-1P-Tyr15 was detected in the nucleus of interphase blastomeres but was absent from the nuclei of late prophase blastomeres (Figure 8B), suggesting the presence of active CDK-1 in prophase blastomeres of wild-type N2 embryos (100%, n = 20; Figure 8B). We observed a similar pattern of CDK-1P-Tyr15 in the npp-16(ok1839) embryos exposed to normoxia (100%, n = 18; Figure 8B); thus, deletion of npp-16 did not result in abnormal regulation of CDK-1 in a normoxic environment. In wild-type embryos exposed to anoxia, however, CDK-1P-Tyr15 was detected in the nuclei of arrested prophase blastomeres (81%, n = 26), indicating the presence of inactive CDK-1 (Figure 8B). The protein recognized by anti-CDK-1P-Tyr15 is throughout the nucleus, and thus it is not clear whether the inactive form is also localized to the chromosomes in a similar manner as the CDK-1 detected by anti-CDK-1 (Figure 8B, arrowhead). Note that we do not observe identical overlap between the anti-CDK-1 and anti- CDK-1P-Tyr15 in anoxic prophase blastomeres of wild-type embryos, suggesting that various isoforms of CDK-1 are located within the arrested prophase nucleus. In marked contrast to the wild-type arrested prophase blastomeres, CDK-1P-Tyr15 was not detectable (79%, n = 29) in prophase blastomeres of npp-16(ok1839) embryos exposed to anoxia, suggesting that CDK-1 is in an active state in these blastomeres (Figure 8B, arrowhead). In these experiments we only assayed prophase blastomeres of embryos in which CDK-1P-Tyr15 was detectable in interphase blastomeres to exclude the possibility that lack of CDK-1P-Tyr15 detection was due to insufficient staining and to ensure that the antibody indeed entered into the embryo. Furthermore, we restricted the analysis to the few prophase blastomeres that are observed in npp-16(ok1839) embryos under anoxia conditions and excluded from the analysis nuclei with the abnormal morphology (as seen in Figure 3). These data show for the first time that CDK-1 localization and posttranslational modifications are altered in arrested prophase blastomeres and suggests that the activity state of CDK-1 plays a role in anoxia-induced prophase arrest and that CDK-1 activity is misregulated in the npp-16(ok1839) prophase blastomeres exposed to anoxia.
We conducted cellular and genetic analysis of prophase cells exposed to anoxia to gain a greater understanding of mechanisms regulating arrest of prophase cells. Using time-lapse microscopy, we found that chromosomes associate with the inner nuclear periphery in anoxia-induced arrested prophase blastomeres. Others have shown that chromosomes associate with the nuclear periphery in Drosophila prophase cells exposed to anoxia (Foe and Alberts, 1985 ), indicating that this response to oxygen deprivation is conserved. We found that oocytes, of anoxia exposed adults, aged adults, or adults that have dysfunctional sperm (fog-2(q71) mutant), contain bivalent chromosomes associated with the nuclear periphery. However, we did not observe this in the oocytes of starved adults. Others have shown that brief, acute, and chronic starvation induces a variety of cellular and behavioral responses. However, the specific response to starvation often depends on the stage at which starvation was induced and how long the animals were exposed to starvation. Known germline responses to starvation include an increase in programmed cell death, arrest of germ cell proliferation, and reproductive diapause. Recently, it was shown that after 5 d of starvation L4 larvae will either arrest as L4 larvae, as adults with the germline in reproductive diapause or become adults in which embryos in the uterus develop and cause death of the adult due to vivipary (bagging out; Angelo and Van Gilst, 2009 ). In our experiment we found that many of the adults were holding embryos in the uterus after 12 h without a food source, and longer periods of starvation led to the bagging out phenotype (data not shown), indicating that the lack of food was indeed inducing a physiological response. However, it may be of interest to determine if a specific starvation regimen does induce chromosome association with the inner nuclear periphery in the oocyte. Embryos and oocytes exposed to the electron transport inhibitor sodium azide contain nuclei with chromosomes associated with the nuclear periphery. However, unlike anoxia exposure, blastomeres exposed for longer periods to sodium azide led to abnormal nuclei suggesting that anoxia-induced arrest is likely not just due to inhibition of electron transport activity. Together, these results demonstrate that chromosome association with the nuclear periphery occurs in both mitotic and meiotic cells and may be a characteristic of arrested or quiescent cells rather than stressed cells.
The relevance of chromosome association with the nuclear periphery, in anoxia-exposed blastomeres or oocytes, is not understood. Many studies have shown that in interphase cells, transcriptional regulation and silencing of genes is associated with the localization of chromatin to the nuclear periphery (Shaklai et al., 2007 ). In Drosophila and yeast it has been shown that chromosomal loci at the nuclear periphery interact with nuclear pores and that nuclear pores can interact with both transcriptionally inactive and active loci (Casolari et al., 2004 ; Akhtar and Gasser, 2007 ). Several nucleoporins as well as nucleoporin-associated proteins (MLP1, MLP2) are required for full repression of specific loci, indicating that NPC has a role in gene regulation (Galy et al., 2000 ; Feuerbach et al., 2002 ). An alternative reason for prophase chromosome association with the inner nuclear periphery, in arrested cells, may be to maintain genome stability. It has been shown that stabilization of repetitive DNA sequences (rDNA repeats), which are prone to homologous recombination, occurs by interactions between inner nuclear membrane proteins and chromosomal proteins (Mekhail et al., 2008 ). In anoxia-arrested prophase cells and oocytes the chromosomes are highly condensed and thus likely to not be transcriptionally active. Furthermore, we found that the quiescent oocytes in the sperm-dysfunctional fog-2 mutant or sperm-depleted aged hermaphrodite also contain oocytes with bivalent chromosomes associated with the inner nuclear periphery; these quiescent oocytes are viable and have other phenotypes similar to that of oocytes exposed to anoxia (Jud et al., 2008 ). We suggest that the relevance of chromosome association with the nuclear periphery in these arrested cells may not be to regulate gene expression but rather to maintain genome stability and/or chromosome structure in times of arrest. Thus, it may be possible that the mechanisms for maintaining oocytes in an arrested or quiescent state are overlapping, regardless if the arrest is induced by anoxia exposure, inhibition of electron transport activity, sperm depletion, or sperm dysfunction. It will be of interest to determine if anoxia-induced chromosome association with the nuclear periphery is important for genome stability and if this phenomenon is an essential component for arresting prophase cells.
The NPC is a large protein channel associated with the nuclear envelope and made up of nucleoporins. This highly conserved protein structure is important for the transport of macromolecules between the cytoplasm and nucleoplasm. Recently, there is increasing evidence that the nucleoporins are involved with a variety of cellular processes including chromatin organization, gene regulation, and spindle orientation (Schetter et al., 2006 ; Capelson and Hetzer, 2009 ). Furthermore, it has been shown that some nucleoporins (Nup107) not only localize to the NPC but, upon NPC disassembly, relocalize to chromatin during mitosis (Galy et al., 2006 ). Phenotypic analysis of the npp-16(ok1839) embryo exposed to anoxia indicates that npp-16 is required for prophase arrest. This is supported by the finding that there is a decrease in arrested prophase blastomeres and an increase in abnormal prophase blastomeres in npp-16(ok1839) embryos exposed to anoxia. Analysis of the npp-16(ok1839) oocytes suggests that npp-16 is also required for maintenance of oocytes. This is supported by the finding that npp-16(ok1839) hermaphrodites exposed to normal air conditions contain bivalent chromosomes associated with the nuclear periphery. The npp-16(ok1839) hermaphrodites exposed to anoxia have oocytes with chromosomes associated with the nuclear periphery, yet many of these oocytes have an abnormally shaped nucleus. These results suggest that in the oocyte npp-16 is not required for chromosome association with the nuclear periphery but is required for maintenance of nuclear shape in oocytes of animals exposed to anoxia.
Using a fluorescently labeled dextran molecule, we show that the permeability of the npp-16(ok1839) embryo exposed to normoxia or anoxia does not appear abnormal (Figure 6). This indicates that the nuclear abnormalities observed in npp-16(ok1839) embryos exposed to anoxia may be due to cellular defects other than those affecting nuclear transport. The C. elegans NPP-16 has homology to the mammalian Nup50. The relative location of Nup50 is peripheral and at the nucleoplasm side of the NPC. Other than a role in nuclear transport, additional functions involving Nup50 are not known. However, using a two-hybrid analysis NUP50 was shown to interact with p27kip1, a CDK/cyclin inhibitor, suggesting a role of NUP50 with cell cycle checkpoints (Sherr and Roberts, 1999 ; Smitherman et al., 2000 ).
We used antibodies to detect CDK-1 and the phospho-specific antibody (CDK-1P-Tyr15) to detect the inactive form of CDK-1 in wild-type and npp-16(ok1839) embryos exposed to normoxia or anoxia (Hachet et al., 2007 ). Interestingly, we found that anti-CDK-1 was detected on or near chromosomes of prophase blastomeres exposed to anoxia. The relevance and activation state of CDK-1 associated with/near chromosomes in anoxic prophase blastomeres is not understood. Others have shown, using mammalian cell culture, that cyclin B1 localizes to chromatin, centrosomes and kinetochore during mitosis yet localization of cyclin B1 to chromatin is independent of CDK1 binding (Bentley et al., 2007 ). It will be of interest to determine if other cell cycle–regulated proteins are localized to chromatin in arrested prophase blastomeres. The lack of CDK-1P-Tyr15 detection in prophase blastomeres indicates that the active form is present; this is the case for late prophase blastomeres of developing wild-type and npp-16(ok1839) embryos. Yet, the inactive form of CDK-1 (CDK-1P-Tyr15) was detected in the arrested prophase blastomeres of wild-type embryos exposed to anoxia. Therefore, a potential mechanism to arrest prophase blastomeres in response to anoxia likely involves the inactivation of CDK-1. This model is possible given that CDK-1 is known to regulate the phosphorylation of several inner nuclear membrane proteins before NEBD (Heald and McKeon, 1990 ; Margalit et al., 2005 ). Interestingly, the nucleus of prophase blastomeres of npp-16(ok1839) embryos exposed to anoxia contained the active form of CDK-1 (CDK-1P-Tyr15 was not detected in). Thus, it is plausible that in anoxia-exposed npp-16(ok1839) prophase blastomeres the activated CDK-1 promotes NEBD, and thus the cells are unable to maintain prophase arrest. This interpretation is further supported by our finding that there is a decrease in the number of arrested prophase blastomeres in npp-16(ok1839) embryos exposed to anoxia. However, further studies will need to determine if the active form of CDK-1 in prophase blastomeres of npp-16(ok1839) embryos exposed to anoxia is the cause or the consequence of the abnormalities observed in the nuclei of these embryos. Additional studies are needed to elucidate the detailed mechanistic relationship between NPP-16 function and CDK-1 regulation in the context of anoxia-induced prophase arrest. In conclusion, our results support the idea that a natural stressor such as oxygen deprivation, regulates cell cycle arrest and that the nucleoporin NPP-16 and cell cycle regulator CDK-1 may have a role in a previously uncharacterized prophase checkpoint in the C. elegans embryo.
We especially thank Dr. Sepp Kohlwein for helpful and insightful comments on the manuscript. We thank members of the Padilla lab and the UNT DPGR group for insightful discussions regarding this research and the Caenorhabditis elegans Genetics Stock Center, the Oklahoma Knockout Consortium, and Dr. Askjaer for strains. We acknowledge and thank Dr. Lon Turnbull for technical assistance with confocal microscopy, and Dr. Landon Moore (University of Southern Indiana), Dr. Golden, and Dr. Schumacher for antibodies. This work was supported by Grant R01-GM069419 from the National Institutes of Health, National Institute of General Medical Sciences to P.A.P.
This article was published online ahead of print in MBC in Press (http://www.molbiolcell.org/cgi/doi/10.1091/mbc.E09-09-0787) on January 6, 2010.